Invertebrate Biology 120(3): 237-247.
C 2001 American Microscopical Society, Inc.
Reproduction and larval development of Poilydora robi (Polychaeta: Spionidae),
an obligate
commensal
of hermit crabs from the Philippines
Jason D. Williamsa
Department of Biological Sciences, 100 Flagg Road, University of Rhode Island, Kingston, RI 02881-0816, USA
Abstract. The reproduction of a recenrtly described spionid polychaete, Polydora robi, is examined from the Philippines. Adults inhabit a burrow in the apex of gastropod shells occupied
by hermit crabs. Females were found to deposit broods of 18-94 egg capsules in the summer
(June-August) and winter months (January-March) sampled. Paired or single egg capsules are
attached by stalks to the inside wall of the burrow and contain 40-106 eggs which average 97
pjm in diameter. The total number of eggs per brood ranges from 941-8761 eggs and is positively correlated with the total number of segments and length of female worms. Adults of P.
robi are polytelic, producing <9 successive broods over a 3-month period; a mean of 6.7 d
was exhibited between broods in the laboratory. Femnalesutilize sperm stored in the seminal
receptacles during successive spawnings. Development occurs within egg capsules until the 3segment stage, at which time the planktotrophic larvae are released. Juveniles of -20 segments
are competent to settle on gastropod shells inhabited by hermit crabs. Members of P. robi are
relatively fecund, semicontinuous breeders; the life-cycle in this species is similar to the only
other known obligate polydorid commensal of hermit crabs.
Additionalkey words: Annelida, fecundity,life-cycle, symbiosis
Polychaete worms exhibit a complex variety of life
histories and reproductive modes. The reproductive biology of these worms has been reviewed with particular reference to endocrinology (Clark 1965), physiology (Bentley & Pacey 1992), and reproductive mode
(Schroeder & Hermans 1975; Wilson 1991; Giangrande 1997). Recently, the reproduction and larval
development of the Spionidae and related polychaete
families have been reviewed (Blake & Arnofsky
1999). The life history, fecundity, and population ecology of soft-bottom spionids have been studied in significant detail (Levin 1984; Levin & Huggett 1990;
Zajac 1991; Levin & Bridges 1994; Bridges & Heppel
1996). While larval development of shell-burrowing
polydorids is well documented, studies of their fecundity are lacking due in part to the difficulty in extracting specimens from their burrows. Among the -70
described species of Dipolydora and Polydora, the fecundity of individuals in 12 species has been investi-
gated (Wilson 1928; Campbell 1955; Dorsett 1961a;
Blake 1969; Radashevsky 1986, 1988, 1989. 1994;
Sato-Okoshi et al. 1990; Lewis 1998; MacKay & Giba Present Address:
Departmentof Biology, Hofstra Univer-
sity, Hempstead, NY 11549-1140, USA.
Email: biojdw@hofstra.edu
son 1999; Williams & Radashevsky 1999). Of these
species, only one has been investigated from the tropics; the rest are restricted to temperate regions. The
purpose of the present investigation is to provide data
on the reproduction and larval development of Polydora robi, a recently described burrowing polydorid
from the Indo-West Pacific (Williams 2000).
M embers of P. robi are obligate commensals of hermit crabs, found only in the apex of gastropod shells
occupied by host crabs (Williams 2000). These burrows extend from a hole in the apex to an opening on
the columella along the upper body whorls. Adults are
known to ingest the embryos attached to the pleopods
of host hermit crabs and the worms are found in 2.535% of gastropod shells among sites in the Philippines. The morphology, ecology, and feeding behavior
in P. robi have been documented (Williams 2000). The
present report provides data on fecundity and larval
development to the juvenile stage in P. robi adults
isolated in glass capillary tubes in the laboratory and
from field collected specimens. Life history in the species is discussed with reference to its association with
hermit crabs.
Methods
Hermit crabs inhabiting gastropod shells were collected by hand, shallow subtidally (<5 m) in coral reef
238
areas from six provinces of the Philippines (Bataan,
Batangas, Oriental Mindoro, Aklan, Palawan, and
Cebu) from June to August 1997 and January to April
1999. Hermit crabs were either fixed en masse in the
field (relaxation in 3% MgC12 followed by fixation in
10% formalin-seawater solution) or transported to the
laboratory and isolated in divided plastic boxes until
examination. Isolated hermit crabs were maintained in
aerated unfiltered seawater. Hermit crabs were removed by cracking the gastropod shells in a mortar
and pestle constructed of 60-mm galvanized-steel pipe.
After exposing the burrows of P. robi individuals
by cracking the shells, the worms were forced from
their tubes by pipetting a stream of seawater through
the opening in the apex. The worms were then placed
in glass capillary tubes (10-25 mm long, open at both
ends; inside diameter between 0.9-1.1 mm). Total
length, palp length, width at segment 7, and total number of segments of the worms were recorded after relaxation in 3% MgC12. The worms were maintained at
room temperature (-27?C) and under ambient light
conditions in 20 ml of artificial seawater (Tropic Marin', salinity -32%o) in plastic petri dishes for the
duration of the observations. The seawater was completely refreshed every 24 h and worms were fed Tetra<' fish food (crushed morsels or baby fish food) or
hermit crab embryos, of the species Calcinus gaimardii (H. MILNE EDWARDS1848), in excess approximate-
Williams
length). All means are reported with standard deviations.
Larval development was documented through the
release of 3-segment larvae from egg capsules. The
timing of egg production, larval development, and larval release was followed by examination of the worms
at -12 h intervals. Sketches of egg capsules, larvae,
and recently metamorphosed juveniles relaxed in 3%
MgCl2 in seawater were completed using a compound
microscope with drawing tube attachment. These
sketches were scanned into a Macintosh' computer
and images were prepared using the programs Adobe
Photoshop'T@and Adobe Illustrator?. Three-segment
larvae relaxed in 3% MgCl2 in seawater, fixed in 10%
formalin-seawater solution and stored in 70% ethyl alcohol, were prepared for scanning electron microscopy
(SEM). The specimens were dehydrated in an ascending ethyl-alcohol series followed by 4 changes of
100% ethanol. Dehydration was achieved with Peldri
II (Ted Pella, Inc.) by placing the specimens into a 1:
1 mixture of 100% ethyl alcohol and Peldri II for 1 h
at 34?C. The specimens were transferred to 100%
Peldri II for 3 h and then placed in a cool water bath
and allowed to sublime overnight. Dried specimens
were mounted on a copper stub, coated with gold-palladium mixture, and viewed in a JEOL 1200EX SEM.
Results
ly every 48 h.
After isolation, the worms were monitored for the Field collected specimens
production of egg capsule strings between 4 February
Individuals of P. robi ranged in size from 3.0-41.0
1999 and 9 April 1999. For removal of egg capsule mm in length (mean=16.4 + 9.5, n=40) for 24-171
strings, the capillary tubes containing worms and cap- segments (mean =75 ? 25, n= 111). Developing gamsules were immersed in 3% MgCl2 for 1 min followed etes were found in a total of 70 individuals, of these
by a stream of seawater applied via pipette through the 33 had ova in the coelom, 22 had sperm, and 15 intube ending containing the posterior end of the worm. dividuals had both stored sperm and eggs visible. Ova
The number of capsules per string were recorded and were present from segments 17-38 to 28-127 with a
the number of eggs per capsule were counted in a sub- mean number of 24.2 + 17.5
(n=48) segments conset (n=3-18) of capsules. Total eggs per brood were taining ova (Fig. 1A). Sperm were present in segments
determined by multiplying the mean number of eggs 15-26 to 21-110 with a mean number of 38.1 ? 20.9
per capsule by the number of capsules per string. (n=37) containing sperm (Fig. IB). The total number
Worms were returned to the capillary tubes and suc- of gametogenic segments was positively correlated
cessive broods were enumerated as described above; with the total number of segments of the worms (Fig.
egg capsules typically were removed from the capil- 1A, B). Egg capsules were found in all months exlary tubes -<12 h of deposition. Data on the brood size, amined (June-August 1997 and January-March 1999);
number of ovigerous segments, and total number of 8-80 capsules (mean= 27.1 ? 19.2; n =14) were found
segments were also recorded from specimens pre- joined in strings on the inside of shells from the field.
served in the field prior to examination. Gametogenic
segments were counted using a compound microscope;
Reproduction in the laboratory
developing eggs appeared yellow in the body coelom;
Isolated specimens of P. robi produced 1-9 broods
sperm appeared white. Regression analysis was used
to examine the relationship between measures of fe- during the observational period (Table 1). The number
cundity (mean number of capsules per string and total of egg capsules per string did not vary significantly
eggs per brood) and female size (total segments and from their mean over the spawnings, examined in 12
Reproduction and development in Polydora robi
A
A
10string
y =0.68x
- 34.02 r = 0o68
100-
~~~>
80
i
60--
wS"~~~~ ^^~~~total
3^~~~~
>^~~~~941--8761
/
/
/
/total
rnPy~~92C,
O OE
,at
??>Sa~~~~~
_vC~
i-ni5
ra-/0^
~tained
n1
20
r1_M
20
^T Ln~
njtOTju T^~~~~after
0
1
1
1
1
1
1
40
60 80 100 120 140 160 180
s=
40-
Total segments
B
239
ranged from 13.8-93.6 (Table 1) and was positive]ly correlated with both the number of segments
and body length of the worms (Fig. 2A, B). The mean
number of eggs per capsule ranged from 40-106 (Ta~~~~~~~~~ble
1) and was not significantly correlated with either
the total segments or body length of the worms. The
number of eggs produced per brood ranged from
and was positively correlated with both the
segments and body lengths of the worms (Fig.
D). Worms containing sperm in seminal receptacles prior to isolation appeared to be depleted of sperm
the end of the experiment, however all broods confertilized eggs. In 7 cases the worms were found
to have ingested all or part of the egg capsule strings
deposition; these instances may have represented
premratureegg capsule dislodgment from the capillary
tube followed by female ingestion.
Larval development
Females deposited egg capsules in paired or single
attached to the burrow wall by 2 stalks (Fig.
rows,
r2=0.66
y=0.80x-25.36
O
3A). Females usually deposited egg capsules during
the night (91%, n=58), although on 5 occasions the
/
80g
females deposited the egg capsules during the day. Fe,^~~~~~~
y~~/
/
were able to deposit a complete egg capsule
ce^~~~~~ ^~~males
3 60within 1 h. No unfertilized or nurse eggs were
,/string
*"
tfound
and all eggs developed into larvae, which were
/
?released
at the 3-segment stage. These larvae were re,^~~~~~
40leased in 4.6-7.5 d (mean= 5.8 + 0.6, n= 19) and time
between spawnings ranged from 5.5-8.6 d (mean=6.7
g3^ ~ pfflflt~ nz
u
I=2EI
?0.7,n=19).
I0~~~~~~+
20Eggs early in development were circular and had a
E~
3mean
diameter of 97.0 + 6.0 pim (n=50), with a white
El
?to
light yellow/orange color (Fig. 3A). The protro0
X
X
I
,
i
chophore had a rounded anterior end, a small ciliated
20
40
60
80
100 120 140
mouth vestibule, paired ventro-lateral ciliary patches,
and the center was composed of a large yolk area (Fig.
Total segments
3B). At -~3 d the protrochophores measured 128 + 7.0
Fig. 1. Relationshipbetween gametogenic segments versus pm (n=15). Later in development, the larvae postotal segments in individualsof P. robi collected fiom the sessed the cilia of the telotroch and prototroch, two
Philippines in June-August 1997 and January-April1999. kidney-shaped eyespots anterior to the prototroch, and
A. Number of segments containing ova versus totlal seg- yolk center composed of large, irregularly shaped macments. B. Number of segments containingsperm (develop-romeres (Fig. 3C). In 4 d the early
3-segment larvae
ing in nephridiaof males or storedin seminalreceptaclesof were 208
18 (n=8) in length (Fig 4A) The larvae
versus
females)
versustotal
total segments.
females)
segmentsi.^o-i
i
/
ohad 2 kidney-shaped
eyespots and 3 segments with
developing larval spines; only the first set of spines
of the worms producing multiple broods (X2:=0.11protruded through the cuticle. Small ventro-lateral cil7.08, p-<.35-.83, df=l-8). Two worms exhibited a iary patches remained and the yolky macromeres were
reduced in size and were approximately circular in
significant difference (X2=13.78, p-<.03, df=6;
X28.84, p-<.03, df=3); the initial broods of these shape (Fig. 4A).
worms were 27 and 14 egg capsules less than the
In 5 d, the mid- to late-3-segment larvae were 279
mean; additional spawnings of these worms dclidnot + 18 Jm (n=9) and were competent to swim (Fig.
differ significantly. The mean number of capsules per 4B). Two sets of eyespots, a round median pair and a
100
240
Williams
Table 1. Reproduction in specimens of P. robi isolated in glass capillary tubes between 4 February and 9 April 1999.
Number of spawnings (broods) produced per number of days isolated. Number of segments (Seg) and length of the worms
compared to mean number of capsules per string, maximum number of capsules (Max), and mean number of eggs per
capsule. The estimated number of eggs per brood is determined by the product of the mean capsules per string and mean
eggs per capsule.
Worm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Seg
77
80
100
97
62
37
62
71
37
118
171
68
83
105
104
Length
(mm)
Days
isolated
13.3
13.2
16.45
12.3
6.8
6.35
10.95
66
66
54
55
52
52
52
52
52
46
46
38
38
38
38
10.9
5.1
12.15
30.4
8.2
17.2
16.65
14.65
Broods
9
4
7
4
6
4
7
6
2
7
7
Capsules per string
Mean + SD (n)
32
32.8
40.6
39.8
25
? 4.4 (9)
?
?
?
?
13.8 ?
25.8
29.2
25.5
62.6
93.6
?
?
+
+
?
6 (4)
6.2 (7)
5.6 (4)
3.3 (6)
3.9 (4)
4.7 (7)
6.4 (6)
6.4 (2)
12 (7)
6.6 (7)
1
18 (1)
2
5
5
40.5 ? 2.1 (2)
38.8 + 5.3 (5)
38 ? 10.6 (5)
kidney-shaped configuration of 2 pairs of lateral eyespots were observed; 2 tactile cilia were present on the
head (Fig. 4B). The prototroch extended from approximately the mouth vestibule to the lateral eyespots. The
telotroch contained a dorsal gap; nototrochs were present on segments 3 and the developing segment 4. Segment 3 projected laterally and each side contained a
short curved seta (Fig. 4C), -2 pjm in length. The yolk
had been partially depleted and the gut was beginning
to form at this stage; no larval pigmentation was observed.
Fourteen juveniles of P. robi were observed in the
apex of hermit crab shells; additional juveniles were
found together with large females in the apex. Those
shells inhabited by juveniles only did not contain a
hole in the apex as observed for adult worms. Juveniles were composed of 24-27 segments and measured
-1660 pjm in length and -230 Jim in width at segment 7. The palps were short, extending back to segments 5-6. Distinct eyespots were no longer present,
but pigmentation was present on the prostomium between the palps; in some individuals, irregular patches
were found on the dorsal side of the middle segments
(Fig. 5A). The caruncle was short, the triangular occipital tentacle found in adults had not yet developed.
Nototrochs were found on segments 2-3, 7, and posterior segments.
Notosetae in the posterior l/3rd of the body were
found in small bundles of fine needle-like spines pro-
Max
35
37
45
45
29
18
30
35
30
71
98
18
42
43
48
Eggs per capsule
Mean ? SD (n)
82.9
40.0
77.8
46.5
50.1
74.7
58.6
52.0
74
93.6
52.3
105.9
50.2
74.0
Estimated
eggs
per brood
13.1 (7)
10.5 (20)
? 3.9 (9)
? 1.3 (4)
?+ 10.1 (33)
2653
1312
3159
1851
1253
?+ 10.9 (25)
1927
1711
1326
4632
8761
941
4289
1948
2812
?
?
4 (12)
? 3.8 (8)
?+ 20.8 (25)
? 26.6 (39)
? 9.7 (4)
?+ 36 (19)
? 9.6 (21)
? 10.8 (17)
?
truding through the cuticle. The bundles of notosetae
contained -7-10
spines in posterior segments, with
two longer anterior notosetae. Segment 5 was almost
twice as large as segments 4 and 6 and had a slightly
curved row of 3-4 major spines. The major spines
were falcate with a lateral obliquely curved flange. A
tear-shaped gland on each side of the fifth segment was
present (Fig. 5A, C). The gland was ventral to the
major spines and emptied via a duct leading to each
side of segment 5. The glands were composed of -2025 tear-shaped lobes (Fig. 5C, D) which contained a
granular secretion. Later in development, the gland
was reduced in size and the lobes of the glands were
no longer visible (Fig. 5A). Three bidentate hooded
hooks began on segments 7 or 8. Up to 6 hooded
hooks were found in middle body segments and were
not accompanied by capillaries. Branchiae began on
segment 7 and extended to the middle body segments.
Notopodia of posterior segments sometimes contained
knob-like structures and non-motile cilia (Fig. 5B).
The pygidium was cuff-shaped with irregular, rounded
knobs surrounding the anus; knobs possessed non-motile cilia (Fig. 5B). Black pigmentation was present on
the pygidium of some specimens. Glandular pouches
were present in segments 7-9.
Discussion
Members of P. robi are polytelic, capable of producing as many as 9 broods during a 2-month period.
241
Reproduction and development in Polydora robi
B
A
IC:
*;
elm
"ON
^1
ft
3
V
CQ
k;
et
0
50
100
150
200
10000
10
15
20
25
30
35
5
10
15
20
25
30
35
D
"C
0
0
5
8000
6000
4000
2000
0
0
50
100
150
200
Total segments
Length (mm)
Fig. 2. Relationship between measures of fecundity (mean number of capsules and total number of eggs per brood) versus
total body segments and length in individuals of P. robi isolated in the laboratory between 4 February and 9 April 1999.
A. Mean number (+ SD) of egg capsules per string versus total segments. B. Mean number (+ SD) of egg capsules per
string versus length. C. Number of eggs per brood versus total segmnents.D. Number of eggs per brood versus length.
This Indo-Pacific worm produces large broods of eggs;
the largest specimens can produce >8000 eggs in a
single brood. Prior reports on the fecundity o-f polydorids, all confined to temperate waters, had shown
that these species produce 200-5800 eggs per brood
(see Sato-Okoshi et al. 1990). Dipolydora armata
(LANGERHANS1880), from the tropical waters of the
West Indies, produces small broods (50-100, eggs)
composed largely of nurse eggs (Lewis 1998) and
shows no evidence of reproductive seasonality. As indicated by Blake & Arnofsky (1999), most spionids
exhibit seasonality in reproduction, producing eggs
during periods of highest water temperature. Seasonal
variation in water temperature of the Philippines is
limited (26-31?C: Cordero 1981; Yap & Gomez 1981)
and broods of P. robi were found deposited in field
samples during all months examined. The species is a
rapid breeder producing successive broods in <'7 d.
Egg capsule production has been documented in P.
cornuta Bosc 1802 (Rice & Reish 1976). The eggs of
P. cornuta exit the body via paired nephridiopores, and
enter two thin mucous tubules attached to the burrow
wall. As the eggs enter the tubules, the mucus expands
and coalesces, creating a single capsule. These capsules become joined in a string, with additional egg
capsules deposited along the length of the female dorsum. P. robi individuals often produce paired capsules,
apparently because the mucous tubules from each nephridiopore do not coalesce. It is not known if this is
due to environmental, behavioral, or morphological
factors, but the paired capsules are produced in the
field as well as in the laboratory.
Individuals of P. robi exhibit a short period of development in egg capsules (-6 d) until 3-segment larvae are released. The larvae develop in the plankton
until settlement; this length of time is unknown but
242
Williams
u/
Fig. 3. Larval development in P. robi. A. Partial egg capsule string, showing paired egg capsules and attachment stalks
connected to a mucous string (arrow). B. Protrochophore larva in ventral view. C. Trochophore larvae in dorsal view. Scale:
A=300 |tm; B=25 Jim; C=50 jim.
Reproduction and development in Polydora robi
243
A
C
Fig. 4. Larval developmentin P. robi. A. Early 3-segment larva in dorsal view. B. Late 3-segment larva in dorsal view.
C. Curved seta of third segment (redrawnfrom SEM micrograph).Scale: A, B=50 pLm;C=2 pm.
larvae of other polydorids often remain in the plankton
for - 1-2 months (e.g., Woodwick 1960; Dorsett
1961a; Hatfield 1965; Day & Blake 1979; Sato-Okoshi
1994). Larvae of P. robi metamorphose at -20 segments, settling on gastropod shells inhabited by hermit
crabs and completing their life cycle. The observation
that some shells lacking a hole in the apex contained
single juveniles, indicates that the juveniles bore from
the inside of the shell. Juveniles may settle on the outer
or inner lip of the aperture and crawl to the apex of
shells occupied by hermit crabs. After movement to
the apex the worms begin to produce a mucous burrow
along the columella and create a hole in the apex. Fig.
6 shows a diagram of the life-cycle in P. robi. The life
span of most polydorid species is -1 yr, although in
P. brevipalpa ZACHS1933, a burrowing associate of
scallops from Japan, individuals have a life span of
2.5 years (Sato-Okoshi et al. 1990).
Similarities are exhibited in reproduction, ecology,
and behavior between members of P. robi and D. commensalis (ANDREWS1891), the only other obligate
commensal polydorid of hermit crabs. Individuals of
D. commensalis produce large broods (-2400 eggs)
and larvae of this species are released at the 3-segment
stage (Hatfield 1965; Blake 1969; Radashevsky 1989).
As documented in P. robi during the present study,
Hatfield (1965) found males and females of D. commensalis closely associated in burrows; two specimens
244
Williams
A
J "
/(/
,,,/,,,
,/,
i/(l/(^(,.ll,
,,,,,,
{
^
C
(
,,
,'l
Fig. 5. Juvenile of P. robi. A. Anterior end of juvenile in dorsal view. B. Posterior end of juvenile in dorsal view. C. Fifthsegment spines and gland of juvenile; arrow indicates duct of gland. D. Tear-shaped lobe of fifth-segment gland. Scale: A,
B=100 Rm; C=25 Jm; D not to scale.
Reproduction
and development
in Polydora
robi
245
Fig. 6. Schematic diagram of the life-cycle in P. robi. Center of figure shows shell of Drupella cornus inhabited by
ovigerous individual of Calcinus gaimardii and a large female of P. robi occupying a burrow in the apex. Egg capsules
(bottom left) are deposited in the burrow and larvae develop in the capsules until the 3-segment stage. The 3-segment
larvae are released from the opening in the apex or shell aperture and develop in the water column. Juveniles at -20-24
segments (bottom right) settle on the shell and move to the apex. Scale=5 mm, for center figure; rest not to scale.
she examined possessed eggs in the coelom and sperm
in the seminal receptacles. Based on these observations, she concluded that copulation takes place in D.
commensalis but did not suggest a mechanism for
sperm transfer (Hatfield 1965). Rice (1978) indicated
that sperm stored in the seminal receptacles of females
may not have resulted from direct copulation. Instead,
spermatophores produced by males could be picked up
and broken by the palps of females; such sperm would
subsequently be transported and stored in the seminal
receptacles. He suggested that sperm transfer and fertilization in this manner would allow for continuous
breeding and alleviate the need for synchronized
spawning (Rice 1978). The present findings support
this hypothesis.
Specimens isolated for reproductive studies at first
exhibited stored sperm which became depleted during
successive spawnings. The stored sperm were used to
fertilize as many as 9 broods, and all eggs from each
brood followed
normal patterns of development.
has
also been reported in P. hermaSperm storage
HANNERZ
1956 and more recently in P. corphroditica
nuta, where specimens isolated in the laboratory were
& Gibson 1999). However, females of P. ciliata (JOHNSTON 1838), isolated in glass capillary tubes, were not
observed to produce egg capsules unless paired with
a male (Dorsett 1961 a). Spermatophores were not produced by isolated worms or observed in any field collected specimens of P. robi, although they are documented in other members of the genus (Rice 1978).
Sex determination appears to be similar to that in D.
commensalis, in which the first worm to settle on hermit crab shells mature as females and subsequent
wormnsdevelop as males (Radashevsky 1989). Transitory hermaphroditism can occur when the female
dies and the largest male becomes a female (Radashevsky, pers. comm.).
The larval morphology in P. robi is similar to that
of other members of the genus Polydora (see Blake &
Arnofsky 1999). However, the 3-segment larvae possess a distinct curved seta on the third segment, which
has not been previously documented. The seta may
serve a similar function as the grasping cilia found in
some polydorids (Wilson 1928; Blake 1969). The fifthsegment glands found in juveniles of P. robi have been
described in other polydorids (Hannerz 1956; Hatfield
capable of storing sperm for at least a month (M:acKay 1965; Radashevsky 1994), but their function remains
246
Williams
poorly known. Hannerz (1956) revealed the duct of
the glands terminate ventral to the major spines in P.
ciliata. In all species noted, the glands become reduced
or absent shortly after juvenile settlement. Hannerz
(1956) considered the glands to be homologous to the
glandular pouches found in segments 8-10 of some
adult polydorids (e.g., Claparede 1870; Fauvel 1927;
Dorsett 1961b). If these glands are homologous, then
as
they may produce acidic mucopolysaccharides
found in P. ciliata (Dorsett 1961b). The distinct termination of the gland duct by the major spines (Fig.
5C) would allow the mucopolysaccharides
to be utilized, in combination with mechanical abrasion, to
burrow in calcareous substrata (Hannerz 1956). In
spite of past research conducted on the burrowing behavior of polydorids, there is still considerable debate
over the exact mechanism for substrate penetration
(see Sato-Okoshi
1999). A detailed histochemical
study of this gland as it develops and is subsequently
lost may provide insight into its function.
Acknowledgments. I thank Drs. James A. Blake and Robert
C. Bullock for their comments on earlier drafts of this work.
I am grateful to Nancy R. Carson for care of specimens of
P. robi in the laboratory during my collection trips. This
work was supported by the Lerner-Gray Fund for Marine
Research (American Museum of Natural History), a Libbie
Hyman Memorial Scholarship (Society for Integrative and
Comparative Biology), and a Graduate Fellowship from the
University of Rhode Island.
References
Bentley MG & Pacey AA 1992. Physiological and environmental control of reproduction in polychaetes. Oceanogr.
Mar. Biol. Ann. Rev. 30: 443-480.
Blake JA 1969. Reproduction and larval development of Polydora from northern New England (Polychaeta: Spionidae). Ophelia 7: 1-63.
Blake JA & Arnofsky PL 1999. Reproduction and larval
development of the spioniform Polychaeta with application to systematics and phylogeny. Hydrobiologia 402:
57-106.
Bridges TD & Heppel S 1996. Fitness consequences of maternal effects in Streblospio benedicti (Annelida: Polychaeta). Am. Zool. 36: 132-146.
Campbell MA 1955. Asexual reproduction and larval development in Polydora tetrabranchia Hartman. Ph.D. Dissertation, Duke University, Durham, North Carolina. 67
pp. + 35 pls.
Claparede E 1870. Les Anndlides Chdtopodes du Golfe de
Naples. Seconde partie. Anndlides sddentaires. Mdm. Soc.
Physique et Histoire Nat. Geneve 20: 1-225.
Clark RB 1965. Endocrinology and the reproductive biology
of polychaetes. Oceanogr. Mar. Biol. Ann. Rev. 3: 211255.
Cordero PA Jr 1981. Eco-morphological observation of the
genus Sargassum in central Philippines, including notes
on their biomass and bed determination. In: Fourth International Coral Reef Symposium, Vol. 2. Gomez ED, Birkeland CE, Buddemeier RW, Johannes RE, Marsh JA Jr.,
& Tsuda RT, eds., pp. 399-409. Marine Sciences Center,
Manila.
Day RL & Blake JA 1979. Reproduction and larval development of Polydora giardi Mesnil (Polychaeta: Spionidae). Biol. Bull. 156: 20-30.
Dorsett DA 1961a. Reproduction and maintenance of Polydora ciliata (Johnst.) at Whitestable. J. Mar. Biol. Assoc.
U.K. 41: 383-396.
1961b. The feeding behaviour of Polydora ciliata
(Johnst.). Tube-building and burrowing. J. Mar. Biol. Assoc. U.K. 41: 577-590.
Fauvel P 1927. Polychetes sddentaires. Addenda aux Errantes, Archiannelides, Myzostomaires. Faune de France 16:
1-494.
Giangrande A 1997. Polychaete reproductive patterns, life
cycles and life histories: an overview. Oceanogr. Mar.
Biol. Ann. Rev. 35: 323-386.
Hannerz L 1956. Larval development of the polychaete families Spionidae Sars, Disomidae Mesnil and Poecilochaetidae n. fam. in the Gullmar Fjord (Sweden). Zool. Bidr.
Upps. 31: 1-204.
Hatfield PA 1965. Polydora commensalis Andrews-larval
development and observations on adults. Biol. Bull. 128:
356-368.
Levin LA 1984. Multiple patterns of development in Streblospio benedicti Webster (Spionidae) from three coasts
of North America. Biol. Bull. 166: 494-508.
Levin LA & Bridges TS 1994. Control and consequences of
alternative developmental modes in a poecilogonous polychaete. Am. Zool. 34: 323-332.
Levin LA & Huggett DV 1990. Implications of alternative
reproductive modes for seasonality and demography in an
estuarine polychaete. Ecology 71: 2191-2208.
Lewis JB 1998. Reproduction, larval development and functional relationships of the burrowing, spionid polychaete
Dipolydora armata with the calcareous hydrozoan Millepora complanata. Mar. Biol. 130: 651-662.
MacKay J & Gibson G 1999. The influence of nurse eggs
on variable larval development in Polydora cornuta (Polychaeta: Spionidae). Inv. Rep. Dev. 35: 167-176.
Radashevsky VI 1986. Reproduction and larval development
of the polychaete Polydora ciliata in the Peter the Great
Bay of the Sea of Japan. Mar. Biol., Vladivostok 6: 3643.
1988. Morphology, ecology, reproduction and larval
development of Polydora uschakovi (Polychaeta, Spionidae) in the Peter the Great Bay of the Sea of Japan. Zool.
Zh. 67: 870-878.
1989. Ecology, sex determination, reproduction and
larval development of the commensal polychaetes Polydora commensalis and Polydora glycymerica in the Japanese Sea. In: Symbiosis in marine animals, Sveshinkov
VA, ed., pp. 137-164. Academy of Science USSR, Moscow.
1994. Life history of a new Polydora species from
Reproduction
and development
in Polydora
robi
the Kurile Islands and evolution of lecithotrophy in polydorid genera (Polychaeta: Spionidae). Ophelia 39: 121136.
Rice SA 1978. Spermatophores and sperm transfer in spionid
polychaetes. Trans. Am. Microsc. Soc. 97: 160-170.
Rice SA & Reish DJ 1976. Egg capsule formation in the
polychaete Polydora ligni: confirmation of an hypothesis.
Bull. South. Calif. Acad. Sci. 75: 285-286.
Sato-Okoshi W 1994. Life history of the polychaete Polydora variegata that bores into the shells of scallops in
northern Japan. Mdm. Mus. Natl. Hist. Nat. 162: 549-558.
1999. Polydorid species (Polychaeta, Spionidae) in
Japan, with descriptions of morphology, ecology and burrow structure. 1. Boring species. J. Mar. Biol. Assoc. U.K.
79: 831-848.
Sato-Okoshi W, Sugawara Y, & Nomura T 1990. Reproduction of the boring polychaete Polydora variegata inhabiting scallops in Abashiri Bay, North Japan. Mar. Biol.
104: 61-66.
Schroeder PC & Hermans CO 1975. Annelida: Polychaeta.
In: Reproduction of marine invertebrates, Vol. 3. Giese
AC & Pearse JS, eds., pp. 1-213. Academic Press, New
York.
Williams JD 2000. A new species of Polydora (Polychaeta:
Spionidae) from the Indo-Pacific and first record of host
247
hermit crab egg predation by a commensal polydorid
worm. Zool. J. Linn. Soc. 129: 537-548.
Williams JD & Radashevsky VI 1999. Morphology, ecology,
and reproduction of a new Polydora species (Polychaeta:
Spionidae) from the east coast of North America. Ophelia
51: 115-127.
Wilson DP 1928. The larvae of Polydora ciliata Johnston
and Polydora hoplura Claparede. J. Mar. Biol. Assoc.
U.K. 14: 122-128.
Wilson WH 1991. Sexual reproductive modes in polychaetes: Classification and diversity. Bull. Mar. Sci. 48:
500-516.
Woocdwick KH 1960. Early larval development of Polydora
nuchalis Woodwick, a spionid polychaete. Pac. Sci. 14:
122-128.
Yap HT & Gomez ED 1981. Growth of Acropora pulchra
(Brook) in Bolinao, Pangasinan, Philippines. In: Fourth
International Coral Reef Symposium, Vol. 2. Gomez ED,
Birkeland CE, Buddemeier RW, Johannes RE, Marsh JA
Jr., & Tsuda RT, eds., pp. 207-213. Marine Sciences Center, Manila.
Zajac RN 1991. Population ecology of Polydora ligni (Polychaeta: Spionidae). I. Seasonal demographic variation
and its potential impact on life history evolution. Mar.
Ecol. Prog. Ser. 77: 207-220.